Profiling CUDA on Tegra K1 (Shield Tablet)

Recently I have struggled a lot to profile a CUDA application on the Shield Tablet. If you were thinking “What the hell would you need a CUDA app for, on a tablet?” I would understand :D. CUDA it’s not for everyday use but can be very powerful.

As of now (Late 2015), the Shield has the most powerful mobile GPU on the market (Tegra Kepler architecture with 192 streaming processors). I decided to evaluate and profile physics algorithms using such architecture.

Reading through documentations, keynotes from GDC, and presentations I found out that is currently not possible to profile a CUDA application from an APK!

Read More

Deploying Assimp Using Visual Studio and Android NDK for Tegra Devices

Hello folks, welcome back to my blog, hope you are ready for a new adventure. This time I promise it is going to be an adventure with the capital A. I’ve been working on a finite element method algorithm using C++ (and later CUDA) to prove that the latest generation of mobile devices (more specifically the Kepler architecture in the Shield Tablet) is capable of running such complex algorithms.

The Shield is shipped with Android Kit-Kat 4.4 thus using C++ or Java and OpenGL ES 2.0 is not a problem…well not just yet đŸ˜€

Setting up the environment is not too difficult too. I used the Tegra Android Development Pack, that installs, all the tools you need to start developing on Android (including extensions for Visual Studio and the whole Eclipse IDE). After a few clicks you have everything up and running.

Read More